music-caption-rewriter

The saved excerpt is a snapshot from when this skill was listed. The external source remains the complete and most current version.

---
name: music-caption-rewriter
description: Turn a brief music description and optional tagged lyrics into a professional MiniMax Music 3 structured caption with Global Metadata, Vocal Details, and a section-aware Arrangement. Use when users ask to enhance a music-generation prompt, preserve lyric-section directives, retrieve a similar style from bundled templates, fuse styles, or produce JSON or JSONL caption output.
---

# Music Caption Rewriter

Transform the user's musical intent into a new, generation-oriented structured caption. Find useful references through progressive disclosure: route to a small style family, compare compact cards, then read only the selected complete templates.

Use natural-language reasoning and local text files only. Do not execute scripts, build a database, calculate embeddings, call external APIs, or scan all 1,000 templates.

## Inputs

Accept:

- `Caption`: required natural-language music description.
- `Lyrics`: optional lyrics containing bracketed section or control tags.
- Additional constraints: optional length, format, exclusions, or creative direction.

Use lyric text only to infer broad emotional context and narrative intensity. Never quote, paraphrase, summarize, or reproduce it. Treat only bracketed tags as executable structural, musical, vocal, or production directives.

## Workflow

Follow these stages in order:

1. Build a private Music Brief from the inputs.
2. Resolve explicit constraints and section-local directives.
3. Read [references/genre-router.md](references/genre-router.md).
4. Read one primary family index and, only when useful, one secondary family index.
5. Select up to three references with distinct roles.
6. Read only the complete template files named by those cards.
7. Design a coherent section-by-section timeline.
8. Render and validate the new caption.

Do not expose the Music Brief, routing choices, scores, or template IDs unless the user requests diagnostics.

## Build the Music Brief

Extract only supported or reasonably inferred values:

- macro genre, subgenres, and cultural or market style
- mood and emotional arc
- approximate tempo, meter, and groove
- vocal presence, gender, register, timbre, and delivery
- core instruments and production texture
- section structure and section-specific changes
- spatial character and explicit exclusions

Classify each value internally as `explicit`, `tagged`, `inferred`, or `unspecified`.

Do not invent a precise key, BPM, vocal gender, melodic interval, or production technique when a broader description is sufficient.

Preserve an explicit instrumental request. Do not add vocals. If vocal presence is unspecified, choose a conservative treatment supported by the user's description and the closest style family.

## Resolve Constraints

Apply this precedence:

1. Explicit user requirements and exclusions.
2. Section-local directives from lyric tags, within that section.
3. Strong implications from the user's Caption.
4. Selected reference characteristics.
5. Conservative musical defaults.

A section tag may change its local arrangement without replacing the song's global genre. Preserve a hard user exclusion when a tag conflicts with it.

When two explicit instructions conflict, prefer the more specific and later instruction if the intent remains clear. Otherwise make the smallest musically coherent compromise.

Never silently reverse an explicit vocal gender, instrumental requirement, tempo limit, required instrument, or prohibited element.

## Route by Progressive Disclosure

Read the genre router first. Choose:

- one primary family for a clear genre request
- one primary and one secondary family for an explicit fusion
- at most two plausible families for an ambiguous genre
- the general pop and ballad family when only mood or imagery is available

Use genre, groove, instrumentation, and cultural context as stronger routing signals than generic adjectives such as `emotional`, `epic`, `dark`, or `modern`.

Read only th